You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Windows Computer Use Skill installed but native connection unavailable after update 26.5.27.31326 #25512
What version of the Codex App are you using (From “About Codex” dialog)?
26.5.27.31326
What subscription do you have?
ChatGPT Plus
What platform is your computer?
Windows 11 x64
What issue are you seeing?
Computer Use Skill is installed and enabled.
Environment:
Taiwan
ChatGPT Plus
Windows 11
Codex 26.5.27.31326
Problem:
Computer Use appears in the Skills marketplace and is enabled.
However every Computer Use action fails.
Example:
Prompt:
Open Notepad
Response:
"I tried to connect to Windows through the Computer Use plugin, but it reported that the native Computer Use connection is unavailable for this turn."
Other errors observed:
native pipe path is unavailable
Computer Use connection is unavailable
Troubleshooting already performed:
Reinstalled Codex multiple times
Restarted Windows
Verified Computer Use Skill is enabled
Verified account is in Taiwan and on ChatGPT Plus
Verified Codex CLI works normally
Additional findings:
Computer Use Skill exists and can be invoked, but the Windows host never connects.
Under:
%LOCALAPPDATA%\OpenAI
only the following folders exist:
Codex
extension
No Computer Use host folders, services, or processes appear.
The issue started after updating to version 26.5.27.31326.
Computer Use previously worked before the update.
What steps can reproduce the bug?
Open Codex Desktop App on Windows 11.
Verify that "Computer Use: Computer Use Skill" is installed and enabled.
Start a new chat.
Send a simple Computer Use request such as:
Open Notepad
Codex invokes the Computer Use skill.
The request immediately fails with:
"I tried to connect to Windows through the Computer Use plugin, but it reported that the native Computer Use connection is unavailable for this turn."
The issue occurs every time and is reproducible 100% of the time.
What is the expected behavior?
When a Computer Use request is issued, Codex should connect to the local Windows Computer Use host and perform the requested action.
For example:
Prompt:
Open Notepad
Expected result:
Windows Notepad opens successfully.
A byte-for-byte copy using System.IO.File.ReadAllBytes/WriteAllBytes succeeded. The cached node_repl.exe SHA-256 then matched the current app package exactly.
Browser, Chrome, and Computer Use plugins were synchronized to 26.707.72221 and reinstalled.
codex doctor and strict config validation both report OK.
Despite the synchronized runtime, matching hashes, enabled flags, app restart, and new threads, the trusted Windows/Node REPL entry point is still not injected.
This suggests two layers:
WindowsApps EFS prevents normal runtime/plugin reconciliation on affected machines.
Even after repairing the local runtime/cache mismatch, Codex Desktop still does not provision the trusted execution bridge into the thread.
A normal manually configured MCP server is not an equivalent workaround because it cannot reproduce the trusted native pipe and Codex turn/session metadata. Please investigate the Desktop tool-injection/host-service provisioning path for current 26.707 Windows builds.
The newer reports here are useful because they show two separate layers that should not be conflated. On affected machines, the Store update can leave the user-side CUA/Node REPL runtime and bundled plugin cache behind the current package, and Application Protected resources can block normal reconciliation. That local package -> runtime/cache mismatch is within WinBridge's scope: it checks package/plugin/runtime versions and hashes, reconstructs user-owned official-resource copies through verified staging, and validates helper paths/state afterward.
However, the 26.707 report also shows an important boundary: after the runtime was synchronized byte-for-byte and hashes matched, the trusted Node REPL/Computer Use execution bridge still was not injected. WinBridge should not claim to repair that host/tool-injection layer. Likewise, if 26.803 has complete current runtime/helper files but Desktop simply never provisions the native server into the session, that is an upstream Desktop runtime issue.
So for users here the tool may help answer and recover the first question — “are my marketplace/cache/runtime/helper files actually current and consistent?” — while deliberately stopping if those layers are healthy and the failure is the trusted bridge itself. It does not bypass feature gates, sandbox/trust policy, or patch the app package.
Activity
github-actions commentedon Jun 1, 2026
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
tangzfc-tt commentedon Jul 15, 2026
Additional reproducible case on the current Windows build.
Environment:
Observed behavior:
Additional local evidence:
This suggests two layers:
A normal manually configured MCP server is not an equivalent workaround because it cannot reproduce the trusted native pipe and Codex turn/session metadata. Please investigate the Desktop tool-injection/host-service provisioning path for current 26.707 Windows builds.
mshirley1964 commentedon Aug 9, 2026
I can reproduce the same underlying failure on a newer Windows desktop build.
Environment
26.803.41515(released August 6, 2026)Observed behavior
Skills 1 → Computer Use.MCP servers 1 → Computer-usesection is completely absent.Troubleshooting completed
Diagnostic feedback ID from the affected Codex session:
919fe425-0f76-7792-b011-8fd3bef5c4c9This suggests the native Computer Use server/helper is not being provisioned or registered at all, rather than starting and then failing to connect.
zemeng5208 commentedon Aug 12, 2026
I maintain WinBridge Recovery, an independent Windows recovery/diagnostic tool for Codex Desktop local plugin/runtime state: https://github.com/zemeng5208/winbridge-recovery
The newer reports here are useful because they show two separate layers that should not be conflated. On affected machines, the Store update can leave the user-side CUA/Node REPL runtime and bundled plugin cache behind the current package, and Application Protected resources can block normal reconciliation. That local package -> runtime/cache mismatch is within WinBridge's scope: it checks package/plugin/runtime versions and hashes, reconstructs user-owned official-resource copies through verified staging, and validates helper paths/state afterward.
However, the 26.707 report also shows an important boundary: after the runtime was synchronized byte-for-byte and hashes matched, the trusted Node REPL/Computer Use execution bridge still was not injected. WinBridge should not claim to repair that host/tool-injection layer. Likewise, if 26.803 has complete current runtime/helper files but Desktop simply never provisions the native server into the session, that is an upstream Desktop runtime issue.
So for users here the tool may help answer and recover the first question — “are my marketplace/cache/runtime/helper files actually current and consistent?” — while deliberately stopping if those layers are healthy and the failure is the trusted bridge itself. It does not bypass feature gates, sandbox/trust policy, or patch the app package.